To study tissue-mediated control of immunity, we have been using 5 different systems. 1. DENDRITIC CELLS FROM VARIOUS LOCATIONS AS ANTIGEN PRESENTING CELLS: We are studying the effect of different environments on the activation of naive and previously activated T cells. we had previously shown that dendritic cells isolated from mesenteric lymph nodes (which drain the gut) tend to induce naive T cells to produce Il4, IL10 and TGFbeta, while dendritic cells grown from bone marrow or isolated from spleen tend to induce interferon gamma (IFNg) and tumor necrosis factor (TNF). We are now studying the effect of offering antigen in different locations in vivo, and of stimulating T cells in various organs in organospecific culture systems. 2. ANTIGENS GIVEN BY DIFFERENT ROUTES GIVE DIFFERENT RESPONSES: we used the protective antigen (PA) of Anthrax as antigen, and gave it either subcutaneously (SQ) or intraperitoneally (IP). when given SQ, the PA immunizes and leads to protection from anthrax toxin challenge. When given IP, it does not. In contrast, the same antigen given with complete adjuvant immunizes by both routes. The responses, however are different in that PA given with adjuvant gives a response with high levels of Antibodies, as seen by ELIZA, but no or few neutralizing antibodies. Nevertheless, the immunized mice are protected from challenge with anthrax toxin. 3. THE GUT: a. THE EFFECT OF ANTIBIOTICS ON THE HEALTH OF INTESTINAL TISSUE. Mice were given a cocktail of antibiotics for several weeks and then jejunum tissue was assessed by microarray analysis. We found that antibiotics cause expression changes in a number of genes expressed by normal gut epithelium. Some of these are indirect effects due to the loss of commensal bacteria. Others, however, are direct effects on jejunal tissue, as the changes were also found in antibiotic-treated germ free mice. These direct effects were mostly concentrated in mitochondrial functions. b. THE EFFECT OF B CELLS ON THE HOMEOSTASIS OF THE GUT. B cells make up a large proportion of the lymphoid cells in the gut. We have found that their absence leads to a defect in fat metabolism and leptin levels. Gene network analysis showed that there are two intertwined gene networks that are inversely correlated one controlling metabolic processes and a second controlling immune processes (mostly IFN dependent). In the absence of B cells the intestinal epithelial cells upregulate their immune network and downregulate the metabolic one. This occurs only in the presence of gut flora, as it does not occur in gnotobiotic mice. Studies in which we transferred the microbiota from B cell KO and WT mice to germ free recipients showed that the genotype of the host, rather than the source of the bacteria was critical. Thus B cells have a strong effect on normal gut function. The B cell effect is mediated by IgA, as IgA KO mice exhibit nearly the same changes in gene expression as B cell KO mice. Further, many of the genes in the metabolic network are regulated by the transcription factor, GATA4. CVID patients and some HIV patients that have metabolic syndrome, have a similar gene expression phenotype to the B cell KO mice, including changes in expression of GATA4-dependent genes.. We conclude that the gut epithelium can upregulate its own immunity when the immune system is deficient, and that the upregulation in immunity is accompanied by a downregulation in metabolic activity. 4. THE MOUTH: a. THE EFFECT OF COMMENSAL ORGANISMS ON NORMAL MOUTH DEVELOPMENT: In order to evaluate the role of oral bacteria in the homeostasis of oral tissues, oral bacterial samples and oral tissues (tongue, mandible, cervical lymph nodes) from germfree vs. conventional mice have been under collection. The RNA samples of oral epithelia obtained from 6 month old mice are under RNA microarray analysis. b. FUSOBACTERIUM NUCLEATUM AS A POTENTIAL VACCINE VECTOR: In order to determine the potential of Fusobacterium nucleatum (a normal human commensal bacterium) as a mucosal vaccine carrier, we are characterizing the adaptive immune responses to F. nucleatum after sublingual administration. Early results show that both IgA and IgG2a antibody responses and Th1/Th17 T cell cytokine profiles are elicited. We are currently working on optimizing the immunization protocol. c. P gingivalis (the bacterium that causes gingivitis) stimulates dendritic cells to produce IL-5 and not IL12p70. E coli, in contrast, stimulate TNF and IL-6 production. We are studying the interactions of these two bacteria and their effects on the stimulation of dendritic cells. We found that P gingivalis does not activate dendritic cells to produce inflammatory cytokines, though it does induce the upregulation of cell surface costimulatory molecules. We are currently deciphering the mechanisms used by the bacterium to mediate its DC-activating activity. Early results show that the intact bacterium does not behave like the isolated LPS. 5. THE PERITONEAL CAVITY We found that the peritoneal cavity harbors a particular subset of NK cells. Although these NK cells share some markers with immature splenic NK cells, and some other markers with NK cells found in the liver, they do not entirely overlap with either population. They produce IFNgamma, and very few other cytokines, and are able to kill standard NK targets. When whole spleen cells are injected into the peritoneal cavity, the T cells are able to leave, but the NK cells do not. When injected intravenously, only a subset of splenic NK cells are able to enter the peritoneal cavity, and these change their phenotype to become typical peritoneal NK cells. Thus the peritoneum seems to preferentially allow the entry and differentiation of a resident population of NK cells that does not exit.